DISCUSSION
In the present study, we found a significant inverse correlation of LL-37 levels in nasal secretions, with the duration of hospitalization and the need for medication use in children with acute bronchiolitis. For the first time, an intriguing association of LL-37 nasal secretion concentration and disease severity is suggested. To the contrary, no correlation was found between nasal secretions β-defensin-2 or serum 25(OH)D levels and the severity of bronchiolitis. Finally, there was no association between AMPs nasal secretions levels and 25(ΟΗ) serum concentration, age, gender, ethnicity, type of delivery or any modifiable factor such as breastfeeding and smoke exposure.
Our findings are consistent with limited existing data examining the association between LL-37 serum levels and the severity of bronchiolitis. Μansbach et al found a significant inverse correlation between serum cathelicidin levels, but not 25(OH)D levels, and the risk of hospitalization >24h in children <2 years old with bronchiolitis. Serum 25(OH)D concentration or other factors like age, gender or ethnicity did not correlate with cathelicidin levels16. A subsequent sizable multicenter cohort study of infants hospitalized with bronchiolitis confirmed these data, showing an inverse association between serum LL-37 levels and disease severity defined by hospital length of stay and intensive care use17. Finally, this inverse relationship between LL-37 levels and ICU admission was shown in another prospective cohort study in infants with Haemophilus-dominant airway microbiota18. In our study, the lack of association of nasal secretion LL-37 levels with intensive care use may be attributed to the small number of children (n=10) admitted to ICU and stringent ICU admission criteria due to restricted bed capacity.
In our study, AMPs levels in nasal secretions did not correlate with gender, age, ethnicity, gestational age, birth weight, type of delivery or smoke exposure. There are few data in the literature concerning factors that could independently impact AMPs levels in biological fluids. No correlation was found between serum cathelicidin levels and race/ethnicity, BMI and body fat percentage in healthy children, but data concerning age, gender or smoke exposure are conflicting19-23. Finally, in a study of 115 infants, cord plasma LL-37 levels were three‐times higher after normal vaginal delivery compared with Caesarean section24.
According to recent in vitro studies, vitamin D induces the production of AMPs in a wide range of tissues25 including bronchial epithelial cells26. However, the question of whether there is an association between vitamin D and AMPs levels in vivo is yet to be answered. In the present study, no correlation was found between 25(OH)D serum concentration and AMPs levels in nasal secretions. Most of existing studies examine the association between 25(OH)D and AMPs serum levels with conflicting results both in healthy subjects and patients19,27-30.
In the present study, serum 25(OH)D levels were not associated with the severity of acute bronchiolitis. These results are consistent with those of a previous study that found no correlation between vitamin D status in infants hospitalized for RSV bronchiolitis and disease severity defined by length of hospitalization, lowest oxygen saturation and disease severity clinical score31. On the contrary, there are studies showing an inverse association between vitamin D levels and the risk of severe bronchiolitis32,33.
Αlthough AMPs seem especially promising therapeutic tools against viral pathogens, there are certain limitations in their clinical use mainly due to their pharmacokinetics, as they seem to have low metabolic stability and undergo degradation by proteolytic enzymes in the systemic circulation. Consequently, the use of novel techniques that improve AMPs properties, such as synthetic analogues formation, have generated increasing interest within the academic and pharmaceutical communities34-36. Currently, although AMPs can be administered by various routes, many researchers consider topical application the only effective way for safe clinical use.
In our study, we showed for the first time an inverse correlation between LL-37 concentration in nasal secretions and severity of bronchiolitis. Our findings could have potentially significant implications if indeed nasal secretion concentration of LL-37 reflect the impact on lung function closer than serum levels. Interestingly, Barlow et al showed that treatment of influenza virus infected mice with nebulised LL-37 reduces disease severity and viral replication, demonstrating LL-37 antiviral effect in vivo10. Future controlled randomized trials of local treatment with AMPs via inhalation hold promise to provide protection against viral lower respiratory tract infections such as acute bronchiolitis.
There are certain limitations in our study. First, as we have no healthy control population, serum 25(OH)D and AMPs levels in nasal secretions as well as the association between them or with other independent factors was examined only in hospitalized infants with bronchiolitis. Second, only a small percentage of children with bronchiolitis require hospitalization, and so the inverse correlation between LL-37 levels and disease severity may not exist in outpatient infants with mild-to-moderate illness. Finally, our study does not establish causality between low LL-37 levels and bronchiolitis severity.
In conclusion, we found a statistically significant association between LL-37 nasal fluid levels and bronchiolitis severity, independently of vitamin D status. As we found no correlation between nasal secretion LL-37 levels with any modifiable factor that could modulate its endogenous production, we hypothesize that LL-37 concentration in the respiratory tract mucous membranes itself may predict the progression of acute bronchiolitis in young children. By inference, if future exogenous application becomes available, there may be benefits regarding the severity of disease and length of hospital stay. Since there is not any effective treatment for acute bronchiolitis yet, the development of novel alternative therapeutics is of great importance and human LL-37 or its synthetic analogues can be very promising novel drug candidates in the future. Therefore, further research is warranted in order to confirm any beneficial effect of increased LL-37 levels in the nasal secretions of infants with acute bronchiolitis and determine its predictive value and potential for clinical applications.